Machine Vision: Seeing is Believing

EE Times Current · 14 May 2021 · 37m

AI in practice: one real use of AI described

What this episode covers

Jeff Bier, president of BDTI and founder of the Edge AI and Vision Alliance, discusses the evolution of machine vision from constrained manufacturing environments to complex natural settings. He explains how the industry shifted from handcrafted algorithms to data-driven deep learning, enabling applications like mask detection and elder care monitoring. Bier also highlights the necessity of sensor fusion, such as combining vision with radar or LiDAR, to overcome environmental limitations like fog or darkness.

AI in practice

Where this conversation describes AI being used to build or deliver a product. Part of the AI in practice tracker.

Engineering & simulation

Deep learning algorithms were used to train models on large datasets of images (e.g., people with and without masks) to solve real-world computer vision problems that were previously too difficult for handcrafted algorithms due to environmental variations.

“And so now we can actually quite well, solve problems like people walking by in the street. Are they wearing masks or not? Or another really cool application that I came across recently? They call it night nurse. It's from a Dutch startup called Kepler Vision. And it's, um, it's for use in like elder care facilities so that if, uh, an elderly patient, for example, falls out of bed.”

Speaker 2 · Hear it at 9:29

Mentioned

Companies
BDTI, Edge AI and Vision Alliance, Kepler Vision, EE Times
Also
Embedded Vision Summit

Other shows on the same ground, closest first.

  • Manufacturing Happy Hour · 35m

    Vision and Imaging 101 with Edmund Optics' Greg Hollows

    Greg Hollows, VP of Product Strategy at Edmund Optics, discusses the evolution of machine vision from the late 1990s to today, highlighting the shift from low-resolution, expensive offline inspection to high-speed, real-time processing enabled by increased bandwidth and computing power. He identifies key growth areas for imaging technology, including autonomous vehicles, agricultural drone sensing, and semiconductor manufacturing, while noting that physics limitations in optics are becoming a bottleneck alongside data processing. Hollows shares career advice drawn from his 20+ years at the company, emphasizing the importance of diversifying project portfolios to mitigate economic risk, as seen during the 2008 recession, and the value of networking at industry events like A3's Vision Show to stay connected to emerging applications.

  • The Manufacturing Executive · 24m

    Custom Machine Vision & Robotics Solutions w/ Jonathan Berte

    Jonathan Berte, Founder of RoboVision, explains how his company uses AI-driven machine vision to solve the labor shortage and complexity challenges in modern manufacturing. He details the shift from manual algorithm creation to automated deep learning, enabling scalable quality control and object detection for clients like Audi. Berte also discusses the strategic trade-offs between cloud and edge computing for industrial applications and the role of flexible robotics in replacing tedious human tasks.

  • OnTrack · 47m

    AI Vision Systems in Manufacturing with Deepview's Eliyahu Davis

    Eliyahu Davis, lead engineer at Deepview AI, explains how his company uses deep learning to automate factory inspection, reducing setup time from 40 hours to minutes. Davis details his transition from aerospace software engineering to hardware, describing how he learned Altium Designer through online resources and mentorship to design the camera's PCB. He discusses the technical challenges of running neural network training on edge devices and the future of foundation models in industrial automation.

  • Embedded.fm · 1h 11m

    161: Magenta Doesn't Exist

    Kat Scott, a senior software engineer at Planet and co-author of 'Practical Computer Vision with SimpleCV', explains the fundamentals of computer vision, emphasizing that it is largely about discarding irrelevant data to find specific signals. She details the challenges of industrial computer vision, noting that consistent lighting and constrained environments are critical for success, and warns that achieving high accuracy in machine learning requires exponentially more data for each decimal place of improvement. Scott also discusses the technical pipeline behind real-time face filters like those on Snapchat, which involve face detection, 3D mesh fitting, and reprojection, and advises engineers to use tools like OpenCV and ROS rather than reinventing the wheel.

  • Over the Air · 39m

    Embedded Systems and IoT Connectivity: Lessons from Automate 2024

    Ryan Prosser, CEO of Very Technology, reports from Automate 2024, interviewing leaders from Machine Metrics, Unit X, Sensory Robotics, On Logic, Light Guide, Elementary, Snowbottics, Acme Manufacturing, Happily, Kongitech, MyTech, Lincode, Wheel Me, and Orbit 3D. The conversation covers AI-driven vision platforms, robotic safety sensors, rugged edge computing, and autonomous mobile robots, highlighting the shift toward edge AI and human-centric automation. Key technical discussions include point cloud generation for safety, modular industrial PCs, and AI-based defect detection for high-speed manufacturing lines.

The summary, chapters and topics are generated from the episode’s audio by this directory, not written by EE Times Current. The episode itself belongs to its publisher.